The dataset contains the concentrations of 11 essential elements (micro-nutrients) analysed in whole bodies of 78 forage species sampled on the continental shelf or in the canyons of the Bay of Biscay, North-East Atlantic. The species encompass jellyfish, crustaceans, cephalopods, cartilaginous and bony fish from coastal to oceanic and deep-sea waters. The elements include two macro-minerals (nitrogen and phosphorous) and nine trace elements (arsenic, cobalt, chromium, copper, iron, manganese, nickel, selenium, and zinc). Most of the specimens were collected during the annual EVHOE fishery campaigns (“Evaluation Halieutique de l'Ouest de l'Europe”; https://doi.org/10.18142/8) conducted each autumn by the “Institut Français de Recherche pour l'Exploitation de la Mer” (Ifremer) on R/V Thalassa, between 2002 and 2008, by benthic trawling with marge vertical opening or by pelagic trawling. Some species were additionally sampled during the same period from opportunistic fishing boats. As far as possible, the size range of these forage species was selected to match published prey sizes for cetacean predators in the Bay of Biscay.All the material was frozen immediately after collection and kept at –20°C until being processed. In the laboratory, whole organisms were briefly thawed. To reduce inter-individual variability, few to hundreds of individuals (depending on the size of species) were grouped for each species (i.e. constitution of pools) and homogenized using a stainless-steel knife mill, carefully rinsed with ultrapure water between each sample. These pools of whole specimens (corresponding to analytical samples) were frozen again –20°C, freeze-dried and reduced into fine powder until further analyses. A total of 115 samples was finally analysed. The process of organisms and samples (brief thawing, homogenization, freeze-dried and reduction into powder) was done within a maximum of two years after at-sea collection, and the samples (powders) were stored in a clean and dry place until analyses. Total element analyses on samples were then all conducted at the same date (in 2016).Total nitrogen (N) concentrations were determined following the Kjeldahl method (AOAC, 1990). Briefly, this method consists in digesting the samples with sulfuric acid to transform all N present into ammonium sulfate. The solution is then alkalized and the resulting ammonia is determined by distillation into a known volume of boric acid, the excess of which (corresponding to the amount of nitrogen in samples) is finally determined by titration.For all other elements, aliquots of samples (~200 mg dry weight of homogenised powder) were digested using a 6:2 (v/v) mixture with nitric acid (HNO3 69%, Trace Metal Grade®, FisherScientific) and hydrochloric acid (HCl, 34%, Trace Metal Grade®, FisherScientific). Acidic digestion of the samples was performed overnight at room temperature and then in a microwave oven (START-D, Milestone). Total concentrations of the micro-mineral phosphorus (P) and of the essential trace elements arsenic (As), cobalt (Co), chromium (Cr), copper (Cu), iron (Fe), manganese (Mn), nickel (Ni), selenium (Se) and zinc (Zn) were determined by inductively coupled plasma atomic emission spectrometry (ICP-OES, Vista-Pro Varian) and/or mass spectrometry (ICP-MS, X Series 2 Thermofisher Scientific). The quality assurance of these elemental analyses relied on blank and internal standard controls, and on the accuracy and reproducibility of data relative to the certified reference materials (CRMs) used in each analytical run. The CRMs used were TORT-3 (lobster hepatopancreas, National Research Council of Canada/NRCC) and DOLT-4 (dogfish liver, NRCC). Blank values were systematically below the detection limits and CRM values concurred with certified concentrations, with recovery rates ranging between 83% and 116% depending on the elements and on the CRM. Limits of quantification (LOQ) were calculated for each sample, depending on the mass of the aliquot analysed. The few concentrations below LOQ were replaced by half of the LOQ for each sample of concern. Specifically, this concerned 19 and 2 samples out of the 115 analysed for Cr and Ni respectively, corresponding to about 15% of values (for Cr) or much less than 15% (for Ni), in which cases this method of replacing non-detects by half the LOQ may be applied for data analysis (U.S. Environmental Protection Agency, 2000).All elemental concentrations given on a dry weight basis can be converted on a wet weight basis according to the average percentages of moisture given for each species.

Sampling during annual EVHOE campaigns:i) during the day, on the shelf and on the slope: benthic trawling with large vertical opening, during 30 min.ii) at nightfall, beyond the shelf/the slope: pelagic trawl, during 1h.Element analyses were funded by the French Ministry in charge of the environment.